Preterm birth is one of the major causes of neonatal morbidity and mortality.1,2 Idiopathic preterm labor is responsible for approximately 30% of preterm birth (birth before 37 weeks of gestation). Diagnosis of preterm labor, when persistent contractions associated with cervical change have set in, is rather straightforward. The clinician's daily dilemma is to differentiate uterine contractions that will lead to cervical change, and ultimately delivery, from physiological uterine activity that will not lead to preterm birth.1–3 Indeed, physiological uterine activity is present throughout pregnancy and increases with gestational age.4 This uterine activity may be detected by patients and lead to repeated consultations. Currently, identification of preterm labor relies essentially on clinical criteria. Uterine activity is determined by the patient's subjective self-surveillance, followed by external uterine activity monitoring without quantification.1,4 Evaluation of cervical change using digital or ultrasound examinations is a measure of the effect of uterine activity.5 Biological tests, such as fetal fibronectin, can also be used as a prognostic factor, although with unimpressive likelihood ratios.6 Currently, there are no reliable methods to predict which patient with uterine activity but no cervical change will go on to deliver prematurely. Although delay in the diagnosis of preterm labor may result in a lower efficacy of tocolytic agents,7,8 tocolysis of all patients exhibiting uterine activity is not free of risks for mother and fetus.8,9
There is an urgent need to develop tools for identifying uterine activation before the occurrence of cervical change. The search for new markers is probably hampered by the limited knowledge regarding the biochemical mechanisms leading from uterine quiescence to uterine activity powerful enough to expel the fetus. However, it seems that changes in myometrial cell organization precede active labor and delivery.10–14 These changes involve modification of movement of ions through the plasma membrane of myometrial cells leading to changes in electrical properties, improvement in cell-to-cell communication, and consequently, action potential propagation. Uterine electrical activity generated by muscle cells has been recorded in different species using electrodes placed directly on the uterine wall or on the abdominal surface.13,15–18 Uterine electromyographic signal has been shown to have a good correlation with intrauterine pressure.13,15–19 Basic electromyographic activity represented by pacesetters followed by action potentials has been characterized.20 Electrical bursts recorded during labor correspond to groups of action potentials that initiate muscle cell contractions.13 Burst characteristics depend on the number of cells recruited, cell synchronization in step with gap junction density, and type of action potentials generated.17–19 Typical changes in burst characteristics have been described during preterm and term delivery: decrease in duration, increase in frequency and amplitude, and shift in bursts’ frequency component from low frequencies during pregnancy to higher frequencies during labor.13,15–19,21 However, the chronology of electromyographic characteristics’ modification has never been studied during preterm or term labor.
We hypothesize that electromyographic activity changes occur progressively before the onset of delivery, reflecting progressive myometrial biochemical modification. Our objective was to test this hypothesis in a widely accepted animal model of preterm labor induced by mifepristone.21 The aim of this study was to evaluate the chronological transformation of different electromyographic characteristics during preterm labor induced by mifepristone in pregnant rats and, particularly, to identify the characteristic with the earliest change, in the anticipation that it might aid in the early diagnosis of preterm labor.
MATERIALS AND METHODS
Timed pregnant Sprague-Dawley rats were purchased from Charles River Laboratories (Wilmington, MA) on day 15 of pregnancy (term = day 22). All animals were given food and water ad libitum and were exposed to 12-hour day-night cycles. The protocol for the animal studies was approved by the Animal Care and Use Committee of the University of Texas Medical Branch. The number of animals in each group (n = 9) was based on prior studies involving electromyography analysis during preterm labor in the rats.18,22,23 Since 1 mg of mifepristone induces delivery within 24 hours in 2 of 3 rats (unpublished data), 15 animals had to be treated to induce preterm delivery in 9 rats.
On day 17 of pregnancy, a median abdominal incision was performed under sterile conditions, and general anesthesia was induced by intraperitoneal injection of 0.1 mL/100 g of a mixture of 15 mg of ketamine (Ketalar; Parke-Davis, Morris Plains, NJ) and 2 mg of xylazine (Gemini; Rugby, Rockville center, NY), and the distal portion of the left uterine horn was exposed. A microtip catheter connected to a transducer (Millar Instruments, Houston, TX) was introduced into the uterine cavity between the membranes and the uterine wall, in the middle portion of the horn, to measure intrauterine pressure. Uterine electromyographic signal was recorded by using a pair of electrodes consisting of Teflon-insulated, multistranded thread wires (A-M Systems, Everett, WA) sutured onto the uterine wall approximately 1 cm apart and at the same location as the intrauterine catheter. Each pair of electrodes was composed of a pickup and a ground electrode connected to a transducer that recorded the potential difference between the 2 electrodes. The transducer was fixed on the inner side of the abdominal muscles. The animals were then allowed to recover for 12 hours before the start of the experiments. During the experiments, the rats were moving freely in individual cages placed on receivers that acquired the radio frequency signals transmitted by the transducer. Signals were captured by each receiver and relayed to a data acquisition system (Data Sciences Inc, Saint Paul, MN), where they were sampled at 10 Hz, analog band-pass filtered from 0.1 to 5.0 Hz, and analog-to-digital converted. After amplification (using MacLab; ADInstruments, Castle Hill, Australia), signals were collected and recorded using an online computer with data acquisition and analysis software (Chart 4.2; ADInstruments). On day 18, the rats were allocated to receive a subcutaneous injection of 1 mg of mifepristone (RU-486; Schering, Berlin, Germany) or 1 mg of vehicle. Intrauterine pressure and uterine electromyography were continuously recorded in the unrestrained animals from 1 hour before treatment and up to 24 hours after treatment. Twenty-four hours after mifepristone injection, rats were killed with carbon dioxide. Initiation of the delivery process was identified by the presence of at least one pup in the vagina.
Since mifepristone does not consistently result in preterm delivery in rats, only animals delivering within 24 hours were included in the analysis because the progressive and synchronized changes leading to preterm birth may not have occurred or may have occurred later in the animals that did not have evidence of initiation of delivery within 24 hours of the mifepristone treatment. Intrauterine pressure and electromyography recordings made over 1-hour periods were analyzed at 1 hour before and 6, 12, 18, 20, 22, and 24 hours after mifepristone or vehicle injection. Quantitative analysis of uterine activity was performed using intrauterine pressure integral activity, uterine electromyography integral activity, and mean amplitude of absolute values over these 1-hour periods. The characteristics were expressed as percentage change from the 1-hour baseline period.
Quantitative characteristics of representative bursts were analyzed in each 1-hour period. Representative bursts were defined as bursts occurring with a simultaneous increase in intrauterine pressure, thereby excluding bursts with electrical artifacts (brief abrupt change in the electrical signal). Each 1-hour period was systematically reviewed from the beginning, and the first 6 representative bursts were selected for analysis. This method focuses the analysis on uterine activity and minimizes nonuterine background noise, a particularly important consideration in nonrestrained and nonanesthetized animals. Only good signal-to-noise ratio recordings were used for analysis of individual burst characteristics.
The power density spectrum of each burst was obtained by using the fast Fourier transform. The analysis time-window was always equal to 25 seconds to standardize the number of data points in the fast Fourier transform. In prior studies, we determined that the duration of bursts varied between 10 seconds during most of the gestation, up to about 1 minute during term labor.18 A 25-second period was chosen as a compromise to minimize the amount of nonbursting activity included in the power density spectrum calculation during gestation and to include the major part of the burst activity during labor. For each burst, the 25-second window was centered relative to the center of the burst for objectivity (Fig. 1).
The power density spectrum curve is a function of frequency and represents the relative contribution of each frequency to the signal. The power density spectrums of the 6 bursts analyzed in each time period were averaged to obtain a mean power density spectrum for the specified time period. The first characteristic analyzed was the power density spectrum energy corresponding to the integral of the amplitudes of the power density spectrum from 0.625 to 5.000 Hz. Low frequencies were excluded from power density power density spectrum energy analysis because they include most of the background electrical noise. The highest peak amplitude and frequency were also studied.
Statistical analysis was performed using the Fisher exact test, paired Student t test, and Kruskal-Wallis nonparametric analysis of variance. P < .05 was used to denote statistical significance.
Fifteen animals were treated with 1 mg of mifepristone and 9 animals received vehicle. Mifepristone induced preterm delivery 24 hours after treatment in 60% of rats (9/15), whereas no rat delivered within 24 hours in the control group (P < .05). The 9 animals that delivered in the mifepristone-treated group formed the preterm labor group used for further analysis. Figure 2 shows typical intrauterine pressure, electromyography, and power density spectrum recordings from one rat that delivered within 24 hours after treatment with mifepristone (A), compared to that of a control rat (B).
Evaluation of global contractile activity using 1-hour integral activity of intrauterine pressure recording is presented in Figure 3. Rats from the preterm labor group exhibited a significantly higher intrauterine pressure integral activity at 22 and 24 hours after treatment compared with the control group: respectively, median (interquartile range) 254.35% (224.18–258.01) versus 70.07% (40.88–143.43) and 215.58% (104.68–357.99) versus 59.46% (35.31–109.29); P < .05. Analysis of uterine electromyography integral activity and electromyography mean were significantly higher in the preterm labor group compared with the control group at 24 hours after treatment: median (interquartile range) for electromyography integral 134.62% (103.17–309.64) versus 80.51% (58.42–97.88); for electromyography mean 134.66% (102.94–309.64) versus 81.93% (58.33–97.98); P < .05 (Fig. 4).
Individual burst analysis was performed on 4 electromyography recordings in each group. Analysis of 6 bursts selected in each 1-hour period showed a significantly higher electromyography power density spectrum energy in the preterm labor group compared with the control group at 22 and 24 hours after treatment: respectively, median (interquartile range) 33.17% (19.37–58.95) versus 221.59% (88.8–411.04) and 64.89% (23.48–108.65) versus 210.88% (148.04–289.21); P < .05 (Fig. 5). Peak amplitudes did do not differ significantly between the preterm labor and control groups at any posttreatment time period.
Peak frequencies were significantly different between the preterm labor group and the control group from 12 to 24 hours after treatment (P < .05; Fig. 6). A progressive switch from low (mean ± standard error of the mean 0.745 ± 0.044 Hz) to high (2.537 ± 0.392 Hz) frequencies was found in the preterm labor group, starting as early as 12 hours after treatment (P < .05).
The most direct approach to determining myometrial function would be to measure changes in the force developed by the uterine muscle during contractions. Currently, the methods to measure the mechanical forces involved in uterine contraction are limited. Mechanical consequence of contractions can be quantified by measuring intrauterine pressure. Intrauterine pressure has been widely studied and provides useful information about uterine contractility.19 However, any invasive method has limitations during pregnancy. Uterine contractions are necessarily accompanied by changes in electrical activity generated by myometrial cells. Similar to cardiac electrical activity, uterine electrical activity can be recorded and analyzed noninvasively in humans.13,19
The aim of this study was to identify intrauterine pressure and electromyographic characteristics that exhibit early change after preterm labor induction in the hope of assisting in the early diagnosis of preterm labor. According to our results, the different intrauterine pressure and electromyographic characteristics analyzed can be classified into 2 groups: those that either did not change significantly from baseline or did so late in the active phase of labor and those that changed early during preterm labor. Parameters in the first group included intrauterine pressure integral activity, electromyography integral activity, electromyography mean, and electromyography power density spectrum energy. These characteristics are therefore not very useful for the early detection of the initiation of the preterm labor process.
Peak frequency was the only characteristic that changed early during the process leading to preterm birth. Peak frequency is a qualitative characteristic that represents the changes occurring in the type of electrical activity. Analysis of peak frequency showed a shift from low (around 0.75 Hz) to high (around 2.5 Hz) frequencies in the mifepristone group, starting at 12 hours after induction of preterm labor and several hours before initiation of the expulsion of pups. Peak frequency appeared to be the earliest electromyographic characteristic to be affected during preterm labor induced by mifepristone. Therefore, peak frequency analysis may be a useful marker for the early identification of the initiation of the preterm labor process. This shift from low to high frequencies of peak amplitude has been described during term and preterm delivery in rats and humans.17,18,24 However, frequencies observed in rats are higher than frequencies observed in humans (approximately 0.5 Hz) because of the difference in uterine size.19 Maner et al24 recently demonstrated that peak frequency was a good characteristic for predicting term delivery within 24 hours and preterm delivery within 4 days in pregnant women.
The sequence of modifications of the electrical signal identified is consistent with biochemical changes in the myometrium previously described, especially myometrial cell plasma membrane ion permeability and gap junction expression. Physiologically, electromyography mean and integral are determined by the number of cells recruited, which in turn is directly correlated with gap junction density; whereas the peak frequency is related to the frequency of action potentials in the myometrial cells.25,26 The frequency of action potentials within a burst is a direct measure of the rate of the depolarization/repolarization process in the myometrial cells, a process largely governed by Ca2+ flux across ion channels.12 Modifications in myometrial cell plasma membrane ion channel expression precede modifications of contractile activity observed during term and preterm active labor and delivery.11,12 Calcium ion has been shown to be a major determinant of actin-myosin interaction and action potential in the myometrial cell.12 Voltage-dependent calcium channel subunits have been shown to increase, starting at 8 hours after antiprogestin treatment in day 17 timed-pregnant rats.11 These changes in Ca2+ permeability induce changes in action potential, which in turn modify calcium permeability and cell contractility.12 An increase in the density of gap junctions has been described nearing the active phase of labor. Myometrial gap junction protein (connexin 43) expression increases during term and preterm parturition.27 Connexin 43 expression is negatively regulated by progesterone and positively by mifepristone.28 Gap junction density correlates with action potential propagation and the number of cells contracting.10,27,29 These successive modifications of the ion channels and gap junction expression are in agreement with the identified hierarchy of change in the electromyographic characteristics: first, increase of the peak frequency, followed by increase in electromyography integral and mean.
However, caution is necessary in extrapolating results in animals to humans, particularly when progesterone withdrawal before preterm or labor has never been documented in humans. However, the succession of biochemical modifications occurring in the myometrium during preterm labor is common to human and rats.27,30 Therefore, similar myometrial biochemical modifications should induce similar electrical signal modifications. In support of this hypothesis, a similar shift of power density spectrum peak frequency from low to high peak frequency has been described in humans before term and preterm labor.24
Our results have clinical significance since electromyography can be recorded from the abdominal surface in pregnant women with a good quality signal that correlates well with electromyography recorded from the myometrium. Peak frequency is a particularly interesting characteristic because it does not require basal recording for analysis, is comparable from one subject to another, and does not depend on the electrodes’ positions.13,16,18,19
In conclusion, uterine electromyography analysis, specifically peak frequency of individual burst's power density spectrum, appears to be a promising method for the early detection of preterm labor process initiation.
1. Haram K, Mortensen JHS, Wollen AL. Preterm delivery: an overview. Acta Obstet Gynecol Scand 2003;82:687–704.
2. Iams JD. Prediction and early detection of preterm labor. Obstet Gynecol 2003;101:402–12.
3. McNamara HM. Problems and challenges in the management of preterm labour. BJOG 2003;110(suppl 20):79–85.
4. Iams JD, Newman RB, Thom EA, Goldenberg RL, Mueller-Heubach E, Moawad A, et al. Frequency of uterine contractions and the risk of spontaneous preterm delivery. N Engl J Med 2002;346:250–5.
5. Leitich H, Brunbauer M, Kaider A, Egarter C, Husslein P. Cervical length and dilatation of the internal cervical os detected by vaginal ultrasonography as markers for preterm delivery: a systematic review. Am J Obstet Gynecol 1999;181:1465–72.
6. Leitich H, Egarter C, Kaider A, Hohlagschwandtner M, Berghammer P, Husslein P. Cervicovaginal fetal fibronectin as a marker for preterm delivery: a meta-analysis. Am J Obstet Gynecol 1999;180:1169–76.
7. Amon E, Midkiff C, Winn H, Holcomb W, Shumway J, Artal R. Tocolysis with advanced cervical dilatation. Obstet Gynecol 2000;95:358–62.
8. Fisk NM, Chan J. The case for tocolysis in threatened preterm labour. BJOG 2003;110(suppl 20):98–102.
9. Berkman ND, Thorp JM, Lohr KN, Carey TS, Hartmann KE, Gavin NI, et al. Tocolytic treatment for the management of preterm labor: a review of the evidence. Am J Obstet Gynecol 2003;188:1648–59.
10. Miller SM, Garfield RE, Daniel EE. Improved propagation in myometrium associated with gap junction during parturition. Am J Physiol 1989;256:C130–41.
11. Tezuka N, Ali M, Chwalisz K, Garfield RE. Changes in transcripts encoding calcium channel subunits of rat myometrium during pregnancy. Am J Physiol 1995;269:C1008–17.
12. Sanborn BM. Relationship of ion channel activity to control of myometrial calcium. J Soc Gynecol Investig 2000;7:4–11.
13. Garfield RE, Maul H, Maner W, Fittkow C, Olson G, Shi L, et al. Uterine electromyography and light-induced fluorescence in the management of term and preterm labor. J Soc Gynecol Investig 2002;9:265–75.
14. Lopez Bernal A. Mechanisms of labour: biochemical aspects. BJOG 2003;110 suppl 20:39–45.
15. Hsu HW, Figueroa JP, Honnebier MB, Wentworth R, Nathanielsz PW. Power spectrum analysis of myometrial electromyogram and intrauterine pressure changes in the pregnant rhesus monkey in late gestation. Am J Obstet Gynecol 1989;161:467–73.
16. Mansour S, Devedeux D, Germain G, Marque C, Duchene J. Uterine EMG spectral analysis and relationship to mechanical activity in pregnant monkeys. Med Biol Eng Comput 1996;34:115–21.
17. Buhimschi C, Boyle MB, Garfield R. Electrical activity of the human uterus during pregnancy as recorded from the abdominal surface. Obstet Gynecol 1997;90:102–11.
18. Buhimschi C, Boyle M, Saade G, Robert G. Uterine activity during pregnancy and labor assessed by simultaneous recordings from the myometrium and abdominal surface in rat. Am J Obstet Gynecol 1998;178:811–22.
19. Devedeux D, Marque C, Mansour S, Germain G, Duchene J. Uterine electromyography: a critical review. Am J Obstet Gynecol 1993;169:1636–53.
20. Shafik A. Study of the electromechanical activity of the uterus: experimental study. Clin Exp Obstet Gynecol 1996;23:73–8.
21. Harding R, Poore ER, Bailey A, Thorburn GD, Jansen CA, Nathanielsz PW. Electromyographic activity of the nonpregnant and pregnant sheep uterus. Am J Obstet Gynecol 1982;142:448–57.
22. Buhimschi C, Garfield RE. Uterine contractility as assessed by abdominal surface recording of electromyographic activity in rats during pregnancy. Am J Obstet Gynecol 1996;174:744–53.
23. Buhimschi C, Saade GR, Buhimschi IA, Gokdeniz R, Boyle MB, Garfield RE. Effect of stimulatory and inhibitory drugs on uterine electrical activity measured noninvasively from the abdominal surface of pregnant rats. Am J Obstet Gynecol 2000;183:68–75.
24. Maner WL, Garfield RE, Maul H, Olson G, Saade G. Predicting term and preterm delivery with transabdominal uterine electromyography. Obstet Gynecol 2003;101:1254–60.
25. Demianczuk N, Towell ME, Garfield RE. Myometrial electrophysiologic activity and gap junctions in the pregnant rabbit. Am J Obstet Gynecol 1984;149:485–91.
26. Marshall JM. Regulation of activity in uterine smooth muscle. Physiol Rev 1962; suppl:213–27.
27. Tabb T, Thilander G, Grover A, Hertberg E, Garfield RE. An immunochemical and immunocytologic study of the increase in myometrial gap junctions (and connexin 43) in rats and in humans during pregnancy. Am J Obstet Gynecol 1992;167:559–67.
28. Lye SJ, Nicholson BJ, Mascarenhas M, MacKenzie L, Petrocelli T. Increased expression of connexin-43 in the rat myometrium during labor is associated with an increase in the plasma estrogen:progesterone ratio. Endocrinology 1993;132:2380–6.
29. Blennerhassett MG, Garfield RE. Effect of gap junction number and permeability on intercellular coupling in rat myometrium. Am J Physiol 1991;261:C1001–9.
30. Challis JRG, Matthews SG, Gibb W, Lye SJ. Endocrine and paracrine regulation of birth at term and preterm. Endocr Rev 2000;21:514–50.